An anger superiority effect in the rapid detection of emotional faces has been supported by a variety of researchers using both photographic and schematic faces. The current research attempts to test this effect in a sample of older adults and extend this effect from faces to words. Fifty-four older adults and 42 younger adults were compared on their reaction times to same/different decisions for emotion word stimuli (i.e., happy, sad, angry, neutral) presented in horizontal 3-word triplets. Younger adults detected all word types more quickly than the older adults, and both younger and older adults detected angry words faster than both happy and sad words. An interaction between age and stimuli was observed for percent correct (with the greatest age differences found for sad items) but not for reaction time. These results indicate that both older and younger adults detect angry words faster than happy and sad words. This is consistent with previous research indicating that initial detection of emotional visual stimuli may not be affected by age.Keywords Aging . Emotion . Word ProcessingThe present study sought to replicate and possibly extend previous findings regarding aging and threat detection by using visual words presented horizontally rather than faces. Given that research has indicated a general negativity bias in stimulus detection, it is predicted that angry emotion words will be detected more quickly than happy, sad, and neutral words. Additionally, the current study expects that older adults will show a pattern of threat detection similar to that of younger adults.First reported by Hansen and Hansen (1988), the face-inthe-crowd effect, or the anger superiority effect, asserts that angry faces are detected more quickly than happy faces among a crowd of distracter faces. More specifically, Hansen and Hansen (1988) asserted that faces are preattentively processed for threat features. Since then, several authors have found subjects to display faster reaction times when detecting angry faces than when detecting happy faces among distracter faces (e.g., Fox et al. 2000; Krysko and Rutherford 2009; hman et al. 2001; Williams et al. 2005; Schmidt-Daffy 2011).This effect does not seem to be due to low level processing of specific features as the same results have not been found with inverted faces or faces containing only a mouth in isolation. It may be that threatening faces grab attention more quickly and hold attention for a longer period of time than non-threatening faces (Fox et al. 2000). When using distracter faces that were at the perceived category boundary between happy and angry (i.e., 50 % of subjects in a pilot study rated the face as happy and 50 % rated the face as angry), Krysko and Rutherford (2009) found that subjects detected angry faces faster than happy faces and that, as the number of distracters increased, so did reaction time.Detection of threatening faces is faster and more accurate than that of friendly faces for both neutral and emotional distracters, varying crowd sizes, and upright ...